Rasa

Rasa

by Rasa • • AI Chatbots

No reviews yet
Free plan available View plans

Rasa is a self-hosted conversational AI platform for building chat and voice agents, made for engineering-led teams; Enterprise is sold by quote.

Rasa gives a company the tools to build its own customer-facing AI agents and run them on infrastructure it controls. Rasa Technologies, Inc., an independent company based in San Francisco with research and development in Berlin, develops it. Its story began in 2016 with Rasa NLU, and today's Rasa Platform is built around a dialogue engine called CALM. Teams write business processes as flows, and CALM starts the right flow from its description. An agent therefore works through steps the business has written down, such as the questions a money transfer needs. Rasa Studio is the no-code side of the platform. Its flow builder adds messages, collect steps, custom actions and logic branches, and version history records each edit. Conversation Review replays real conversations beside their details and exports them to CSV. The Analytics dashboard tracks volume, containment, latency and CSAT. The Rasa Pro runtime adds the Enterprise Search Policy, which answers from documents held in a vector store, and the Response Rephraser, which sets the assistant's tone. Agents reach customers through the open source Rasa Chat Widget and the Twilio connector for SMS and WhatsApp. Voice channels and a REST API cover phone lines and custom front ends. Rasa suits enterprises that need to own their AI agents and keep conversation data on their own systems. Banks, telecom operators, healthcare providers and government bodies are typical examples. Engineering-led AI product teams are the core users; engineers work in code, and Studio opens flow editing to the rest of the product team. Developers can start on the Developer Edition, a free licence that runs locally or in production with help from the community forum. Larger deployments move to Enterprise, which adds the full platform and premium support and is licensed through the sales team. Typical jobs are customer support, sales enablement and internal assistants that answer staff questions from company documents. Contact centres put the same agent on the phone through Twilio Voice or AudioCodes VoiceAI Connect. Teams deploy it on Kubernetes with Helm and reach business systems through custom actions and MCP servers. Rasa fits organisations that want their chat and voice agents written, versioned and hosted like the rest of their software.

Pricing

View plans
  • Free Developer Edition

    Free

    • No charge; one pricing state with no monthly or annual option
    • Free Rasa licence usable locally or in production
    • One bot per company
    • Up to 1000 external or 100 internal conversations a month
    • Community support in the Rasa forum
  • Enterprise

    Custom pricing

    • Quote only through sales, no public price
    • Full access to Rasa Platform and Premium Support
    • Large-scale deployment with pre-built enterprise security features
    • Other setups outside both plans are arranged with the sales team

Features

  • Included: Visual conversation flow builder
  • Not included: Live chat handover to a human agent
  • Included: Answers drawn from your own content
  • Included: Website chat widget
  • Included: WhatsApp and SMS channels
  • Not included: Multilingual chat with language detection
  • Not included: Lead capture and qualification
  • Included: Chat transcript export
  • Included: Chatbot analytics dashboard
  • Included: Custom bot persona and tone
  • Not included: Widget branding and white label
  • Not included: Pre-built conversation templates
  • Included: Chatbot REST API
  • Not included: CRM and help desk integrations
  • Not included: Sentiment detection in chat
  • Not included: Chatbot A/B testing
  • Included: Voice channel support
  • Included: Chat data retention controls

Additional Features

  • CALM dialogue engine that starts flows from their descriptions
  • Rasa Studio no-code flow builder
  • Collect, logic, custom action and set slot steps
  • Flow version history in Studio
  • Enterprise Search Policy over a vector store
  • Response Rephraser for assistant personality and tone
  • Open source Rasa Chat Widget over SocketIO
  • Twilio connector for SMS and WhatsApp
  • Voice ready and voice stream channels
  • Twilio Voice, Twilio Media Streams and AudioCodes connectors
  • Rasa Pro REST API and REST channel webhook
  • Studio Conversation Review with CSV export
  • Studio Analytics dashboard with containment, latency and CSAT
  • Analytics pipeline for SQL and BI tools
  • Automatic conversation deletion
  • PII management with retention periods
  • MCP server integration
  • External agents over A2A
  • Self-hosted deployment, locally or in production

Best for

  • Enterprises that must host and own their AI agents
  • Banks, telecom operators, healthcare providers and government bodies
  • Engineering-led AI product teams
  • Organisations that keep conversation data on their own infrastructure
  • Contact centres adding voice agents over Twilio or AudioCodes
  • Developers prototyping one assistant before an enterprise rollout

Use cases

  • Automating customer support conversations
  • Running voice agents on phone lines
  • Answering questions from internal documents with retrieval
  • Serving customers over SMS and WhatsApp
  • Sales enablement assistants
  • Internal assistants for employees
  • Reviewing real conversations and tagging them for follow-up
  • Exporting conversation logs for quality checks
  • Tracking containment, latency and CSAT over time
  • Purging old conversations on a schedule

Screenshots & Videos

Explore Rasa in action

Rasa Homepage With The Agent Builder Hero
View full size
Rasa Homepage With The Agent Builder Hero

User Reviews

Write a Review

No reviews yet

Be the first to share your experience with this software.

Compare with Similar Software

Click to compare side by side with Rasa

Want to claim this brand?

To claim this brand, please contact support.